Joint calibration of Hamiltonian and SPAM parameters

Determine which calibration assumptions are sufficient to jointly learn the Hamiltonian parameters and state-preparation-and-measurement parameters under computational-basis access with experimental SPAM errors.

Background

The paper analyzes known local depolarizing preparation and measurement noise for several translation-invariant and spatially varying IXZ reconstruction results. It does not address general unknown preparation and measurement errors.

The unresolved calibration problem is to identify assumptions under which the Hamiltonian and SPAM parameters are jointly identifiable and efficiently estimable, rather than treating the noise model as known.

References

An important open question is which calibration assumptions are sufficient to learn the Hamiltonian and SPAM parameters jointly~.

— Out-of-control Hamiltonian Learning  (2610.06709 - Gong et al., 5 Oct 2026) in Section 6, Outlook, paragraph “Toward experimental implementation”